com.hankcs.hanlp.classification.features

Class ChiSquareFeatureExtractor

java.lang.Object

com.hankcs.hanlp.classification.features.ChiSquareFeatureExtractor

public class ChiSquareFeatureExtractor
extends Object

特征提取，用来生成FeatureStats并且使用卡方测试选择最佳特征

Field Summary

Fields

Modifier and Type	Field and Description
protected double	chisquareCriticalValue 在P值（拒真错误概率）为0.001时的卡方临界值，用于特征选择算法
protected int	maxSize

Constructor Summary

Constructors

Constructor and Description
ChiSquareFeatureExtractor()

Method Summary

Modifier and Type	Method and Description
Map<Integer,Double>	chi_square(BaseFeatureData stats) 使用卡方非参数校验来执行特征选择 https://nlp.stanford.edu/IR-book/html/htmledition/feature-selectionchi2-feature-selection-1.html
static BaseFeatureData	extractBasicFeatureData(IDataSet dataSet) 生成一个FeatureStats对象，包含一个分类中的所有词语，分类数，实例数。这些统计数据 将用于特征选择算法。
double	getALevel()
double	getChisquareCriticalValue() 获取卡方临界值
ChiSquareFeatureExtractor	setALevel(double aLevel)
void	setChisquareCriticalValue(double chisquareCriticalValue) 设置卡方临界值

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

chisquareCriticalValue

protected double chisquareCriticalValue

在P值（拒真错误概率）为0.001时的卡方临界值，用于特征选择算法

maxSize

protected int maxSize

Constructor Detail

ChiSquareFeatureExtractor

public ChiSquareFeatureExtractor()

Method Detail

extractBasicFeatureData

public static BaseFeatureData extractBasicFeatureData(IDataSet dataSet)

生成一个FeatureStats对象，包含一个分类中的所有词语，分类数，实例数。这些统计数据 将用于特征选择算法。

Parameters:

dataSet -

Returns:

chi_square

public Map<Integer,Double> chi_square(BaseFeatureData stats)

使用卡方非参数校验来执行特征选择
https://nlp.stanford.edu/IR-book/html/htmledition/feature-selectionchi2-feature-selection-1.html

Parameters:

stats -

Returns:

getChisquareCriticalValue

public double getChisquareCriticalValue()

获取卡方临界值

Returns:

setChisquareCriticalValue

public void setChisquareCriticalValue(double chisquareCriticalValue)

设置卡方临界值

Parameters:

chisquareCriticalValue -

setALevel

public ChiSquareFeatureExtractor setALevel(double aLevel)

getALevel

public double getALevel()